YT YouTube2Text

YouTube2Text → Transcripts → I Built An $85M AI Startup in 5 Months (here's my playbook)

I Built An $85M AI Startup in 5 Months (here's my playbook) — Transcript

by Chris Donnelly · 4,281 words · 646 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:00My business just hit $3 million in
0:02annual recurring revenue and got valued
0:03at $85 million just 5 months after we
0:06launched it. In this video, I'm going to
0:07walk you through the exact playbook that
0:09I used to find a great idea, get our
0:11first customers, and build a product
0:13people are desperate to use. Right now,
0:15AI has made it easier than ever to
0:16launch a business and generate revenue,
0:18and that's why I'm going to take you
0:19through my code framework to spot and
0:21validate opportunities so you can build
0:23something of your own. And let me be
0:25clear, this is the fourth time I've
0:26built a business past eight figures
0:28across four different industries and
0:30four completely different products.
0:31Everything I'm about to show you is
0:33repeatable, and these are the seven
0:34steps that I used to launch Searchable
0:36and how you can do the same thing.
0:40In September, I got on a plane and I
0:42flew to San Francisco, and I networked
0:44with dozens of engineers, and I had Zoom
0:46calls with hundreds, and I sat down with
0:48two engineers specifically, which was
0:50Arya and Sam. They had deep technical
0:52expertise, they were obsessive about
0:55product, and were as convinced as I was
0:57that AI search was going to be one of
0:59the most important spaces built in the
1:01next 5 years. You really need to first
1:03understand the three main skills a
1:05founding team requires in order to build
1:07something that can scale. So, number one
1:09would be UX
1:11and design. So, someone who understands
1:14the product. And right now, it's very
1:16easy to vibe code a basic product, but
1:18all this does is lead to a lot of
1:20products that look identical. So, this
1:22person's role therefore is to be [music]
1:24the innovative driver and for them to
1:26take ownership of the whole product
1:28experience and the value that the
1:30customer gets. And yes, they can use
1:32cool code or they can use lovable or
1:34whatever, but they have to really make a
1:36product and design experience that
1:38delivers value to that customer. So,
1:40number two is deep engineering
1:43experience.
1:44And this person takes the final
1:46ownership over the actual build and the
1:48architecture and is responsible for
1:50making the product real. The third skill
1:53you need, truly, is sales and marketing.
1:56And this person is the front man for the
1:58business. They're the owner of
2:00distribution, the story, and they become
2:02the growth engine for the company. So,
2:04within that, you're going to have to
2:05have some overlap, of course. You will
2:07work on the same things. But, this gives
2:09you each a space to own yourselves.
2:11[music]
2:11And between the three of us at
2:12Searchable, all of those bases were
2:14covered. Aryan and Sam on the technical
2:16and product side, and me focused on
2:18leading the business,
2:19having the vision, and focusing on the
2:21distribution. And the real advantage
2:23comes when everyone shares the same
2:25mindset. A strong interest in
2:26technology, a bias towards speed, and
2:29then the ability to use AI to increase
2:32your output. And this isn't a new model.
2:34It's how some of the biggest tech
2:35companies ever have been built. If you
2:37look at Zuckerberg and Facebook, he
2:38started it with his college roommates.
2:40And even when you look at companies like
2:41Apple, you had Steve Jobs and you had
2:44Steve Wozniak. And then later they
2:45brought in Mike to lead the commercial
2:47side of the business. So, neither of
2:48those companies, and many of the
2:50companies that you see growing, don't
2:52start with a single founder. They start
2:54with a shared vision in a small team of
2:56people who want to build a great
2:58company. And if you want to build a
3:00fast-growing company, you will need to
3:02do the same thing. So, you don't need to
3:04be extreme as I was here. You don't need
3:06to fly to San Francisco. Your founding
3:08team could genuinely be in your network,
3:10or even your colleagues.
3:12But, once you've got your founding team
3:13in place, this next phase will decide if
3:15your idea is going to be a big enough
3:17idea to reach a hundred million dollars.
3:24So, the first job of your founding team
3:26is to validate the idea that you've come
3:27up with. So, between you and your
3:29co-founders, you've probably got the
3:31right skill set to build almost
3:32anything. So, don't get tied down to one
3:35vision if there's no proof that it's
3:37actually going to work. So, there are
3:38two parts to this validating the idea
3:41process. The first is assessing whether
3:42the idea is possible, and the second is
3:44working out if people even want it. So,
3:47over the years, I've developed a
3:48framework called code, which
3:50specifically is built to validate
3:52business ideas and help founders to
3:54choose the ones that they should commit
3:56their time to. So, the first part of
3:58code is consumer trends. So, I want you
4:00to study where consumer behavior, where
4:03technology, and demand are shifting
4:05right now. And the aim is to spot what I
4:07call a rising tide. The core reason for
4:09starting SearchBar is that we noticed
4:11that when people are searching for a
4:13product or service, they are no longer
4:15Googling it. They are going to ChatGPT
4:17and trying to find a business or a
4:19service that can help them with their
4:20problem. And so, for us, that was a huge
4:23consumer change, a massive change in how
4:26people search for products, services,
4:28and businesses. You need to read every
4:30industry report, forum threads, and
4:32LinkedIn comment sections. And what
4:34you're looking for is a large number of
4:35people who are experiencing the same
4:37problems at the same time. And when you
4:39find that, you move on to the second
4:41step, which is opportunity. So, finding
4:43a rising tide isn't the hardest part.
4:45The trick is to locate a specific
4:47opportunity within that [music] trend.
4:49So, you're looking for overlap where the
4:51trend is heading and a problem that
4:53hasn't been addressed yet. So, for us,
4:55the answer became clear very quickly. We
4:58knew people were moving towards ChatGPT
4:59[music]
5:00for their answers. So, we knew that what
5:02we had to do was create a specific tool
5:05that made customer acquisition for
5:07businesses possible within ChatGPT and
5:09[music] AI. So, once you know there's an
5:11opportunity, figure out if there's
5:12demand, which is the third step of code.
5:16The way I did this was I spent 6 hours a
5:18day for a month [music] reading
5:20everything I could find about AI search.
5:22Every forum thread, every LinkedIn post,
5:24every comment section, every subreddit
5:26where founders and marketers were
5:28talking about it. And by the end of
5:29that, I didn't need a survey to prove
5:31the demand was there. It was so obvious.
5:34Wherever I looked, people were asking
5:36the same questions constantly, and no
5:38one had a clear answer. So, what you're
5:40looking for is a conversation that's
5:42already happening. If a large community
5:45of target customers already complain
5:47about a problem, then that [music] is
5:48active demand. And when active demand
5:51exists, you just need to insert your
5:53product into the conversation that's
5:54already happening, and that makes
5:56building a business so much easier. One
5:59of the biggest mistakes that founders
6:00make is they say, "There are no
6:02competitors for my product." Whereas in
6:04reality, what I want to see is a
6:06business where there are lots of
6:08previous businesses solving the same
6:10problem, but doing it with traditional
6:12methods. So, for example, SEO tools
6:15exist and build hundreds of millions of
6:17dollars in revenue, and so building an
6:19AEO tool has validation that there is
6:21already budget in that space. The final
6:23question that most founders overlook,
6:25but this will determine the scale you
6:27can grow your business to, is economic
6:29sizing. And this is a very simple
6:31equation. It's total potential customers
6:34times by total customer spend, which
6:37gives you total addressable market. You
6:40want to avoid small markets, as you will
6:42need to completely dominate in those
6:44spaces to make good money. Instead, I
6:46want you to focus on big markets, where
6:48you only need to capture a few percent
6:50to build a big business. So, here's how
6:52the math worked for my last startup that
6:54was in the care home space. First, I
6:57Googled how many care homes there were
6:58in the UK and the US, and Google says
7:0158,000. [music]
7:02Next, I Googled how much does an
7:04individual care home pay per month in
7:07software cost, and I learned they
7:08roughly pay two to $5,000. So, let's
7:12call it $3,000 a month to be
7:13conservative, or $36,000 a year. So, I
7:16then plug it into my economic sizing
7:18formula. 58,000 care homes times $36,000
7:23a year equals a multi-billion dollar
7:25opportunity. So, now that we've been
7:27through each step of code, Aurelia, Sam,
7:30and I were very confident that we had an
7:32idea that people needed and they wanted,
7:35and it was [music] big enough. And as
7:36soon as we knew there was demand for the
7:38idea, we started building. And the thing
7:41that most founders will get completely
7:42backwards is building a product in
7:44silence for months without driving
7:46awareness or testing their idea. And
7:49they'll wonder why the demand is so low
7:51when they finally launch the product. So
7:53instead, I want you to build an audience
7:54before you build the product. It's
7:56exactly what we did at Searchable and is
7:58the single best approach for proving you
8:01have an idea that people actually want.
8:06So the way you're going to want to do
8:07this is with a wait list. A wait list
8:09costs almost nothing to set up, but it
8:11gives you real people with real intent
8:14putting their hand up and giving you
8:16their contact details, which is so
8:18valuable for later down the line when
8:20you launch. [music]
8:21But launching this is going to be the
8:22first test of whether the code
8:24assessment for your business is actually
8:26accurate. So truthfully, if you can't
8:28get 200 people onto a wait list, you're
8:31going to really struggle to sell anyone
8:34later on. So while Arya and Sam and I
8:36were heads down working on the
8:37Searchable product, I was doing
8:39something that, to be fair from the
8:41outside, didn't look like I was building
8:43a product.
8:44>> [music]
8:44>> Every day from about October onwards, I
8:46started putting out content about AI
8:48search, teaching people about it and
8:50explaining how their business could win
8:52in AI search. Importantly, I wasn't
8:55trying to sell our tool. I wasn't
8:56selling hard to the audience. I just
8:58shared genuinely valuable educational
9:01content that people found useful for
9:04their businesses at the time. And at the
9:05bottom of every piece of content, we put
9:07one single line: "Sign up to be the
9:09first to know when our business
9:10launches." So what happens next is a
9:12thing I want to really understand
9:14because it explains why the wait list is
9:16so powerful for a business launch. You
9:19see, if you're trying to sell to people
9:21organically on social media, a good
9:23result, [music]
9:23like a good outcome for you, would be 2
9:26to 5% of people signing up to the
9:28product. But the reality is that is a
9:31terrible conversion rate. If you offer a
9:33wait list with no financial commitment,
9:36the barrier to entry is their [music]
9:37name and their email address, which in
9:392026 people are happily able to give
9:42you. And so, the percentage of people
9:44who sign up is dramatically high. It's
9:4650, 60, sometimes 70% of people. So, by
9:50the time you come around to actually
9:51launching your product, you could have
9:53thousands of people who have put their
9:55hand up and said they want access to
9:57your tool when it launches. So, you're
9:59probably thinking this only works if you
10:01have a large following, but it is simply
10:03not true. The principle works at any
10:05size. [music] Whether you've got 100
10:07followers or 100,000 followers, the
10:09mechanism is identical.
10:11>> [music]
10:11>> You create useful organic content to
10:13educate your ideal customer profile on
10:16the problem that your product actually
10:18solves for them. You give them a reason
10:20to follow you, and then you capture
10:21[music] their intent before you ask them
10:23to spend any money. But this whole
10:25section will be a waste of time and
10:26[music] effort if you don't perfect the
10:28next and most important step. So, in the
10:31early days, it's very easy to get
10:32carried away and build your entire
10:34product. But if you do that, you
10:36honestly can spend six months building
10:38features that nobody has asked for and
10:40nobody will use. So, instead, you need
10:43to do this. You want to build one
10:44feature for one ideal client that solves
10:47one specific problem. That's your MVP,
10:50your minimal viable product. Not a whole
10:52product or a platform, just one thing
10:54that provides value to your customer. I
10:56learned this lesson the hard way and was
10:58actually saved by one of my investors,
11:00Tom Blomfield, who's the founder of
11:01Monzo. We were months into building
11:04Lossie, and he came to us. We asked him
11:07for investment, and he said he wouldn't
11:09invest in the current business, but if
11:11we could launch our MVP of our care home
11:14marketplace in two weeks, he would give
11:16us 100,000 pounds. [music] And so, we
11:18went home that night and thought to
11:20ourselves, we're going to have to strip
11:21this down to the most basic version to
11:24get this thing live and secure this
11:25investment, but it was the best lesson
11:27that we ever learned because we focused
11:29on the one specific part of the product
11:32>> [music]
11:32>> that we knew customers were interested
11:33in and we just built that. And every day
11:36after that, do you know what happened?
11:38Our customers told us what they wanted
11:40us to build next with feedback every
11:42single day. So now if we jump back to
11:44SearchBar, the business I'm building
11:46today, Aryan Sam and I, we were in
11:48London in November and we asked
11:50ourselves the same question. The same
11:53question Tom had asked me years before.
11:54What is the one thing that we can build
11:56right now that proves this idea works?
11:59And the answer was prompt tracking, a
12:01tool that would scan the major AI
12:03platforms, ChatGPT, Gemini, Perplexity,
12:06and show businesses exactly where and
12:09how often they were being mentioned in
12:11the AI answers. So once you're in this
12:13position, I want you to be really honest
12:14with what you and your founders can get
12:16live right now. [music]
12:18Not in two months, not in three months,
12:19but right now. Make it deliberately,
12:22almost aggressively simple.
12:24>> [music]
12:24>> The goal of the MVP isn't to uh like
12:27impress hundreds of people, but it's to
12:30give customers something to try and to
12:32get them to tell you what you should
12:34build next. And if they use what you've
12:36built and they start asking for more,
12:37that is the perfect signal. And you can
12:40use that feedback now to create the rest
12:42of your product roadmap. So now you're
12:44at point where you can effectively
12:46launch your business, but this next step
12:48will define its success and you can't
12:50rush this. So mid-December, Sam and
12:52Aryan I were sitting there and for the
12:54first time we opened SearchBar up to a
12:56larger selection of people from our
12:58waitlist and we gave them an early
13:00[music] access deal, which was a
13:02lifetime discount of 50%. So at this
13:04point it's not so much about the money,
13:06it's about getting early product
13:08champions into the product, using it
13:10every day and giving you feedback every
13:12single day on what they would improve.
13:14So every single day throughout December,
13:16we were collecting feedback. We were
13:18asking what was working, what was
13:19confusing, what was broken, what was
13:21missing, what we could build more of.
13:24And what you're trying to do here is
13:25build a feedback flywheel. So the faster
13:28that the product actually improves, the
13:29more confident your early users become
13:32that this is something that they would
13:33pay for and that they should tell their
13:35friends about this. [music] And the more
13:36confident you become in your own
13:38product. So, to make that flywheel spin
13:41even faster, we then launched a Slack
13:43community. And this became a place where
13:45everyone who was using Searchable could
13:47talk to each [music] other to share what
13:49they were seeing, to ask questions, to
13:50post feedback, and learn from the
13:53product and learn from us about this new
13:55space. So, on top of this, I was
13:57spending, honestly, 10 hours a day on
14:00calls with our current users to
14:02understand what they wanted and what
14:04wasn't working. And then Sam and Aria
14:06would work throughout the night on
14:08fixing the issues and building new parts
14:11of the product.
14:15So, at this stage, Searchable is
14:17improving rapidly. The feedback loop is
14:19in full effect, and our job now is to
14:21reach new customers, to actually scale
14:23the amount of people using the product.
14:25So, I've seen so many technical founders
14:27obsessively focus on the product itself
14:29and almost bolt on marketing afterwards
14:31when it's time to launch. But that puts
14:33so much unnecessary pressure on the
14:35launch week. And a content machine
14:37mitigates that risk and is made up of
14:39three fundamental layers that work
14:41together. So, the first layer is one
14:43that founders either underestimate or
14:44avoid entirely, but it's something I've
14:46seen incredible benefit from over the
14:48past 4 years, which is founder-led
14:50content. Posting consistently on
14:52platforms like LinkedIn builds trust, it
14:54attracts your ideal customer, and it
14:56builds an interested audience. And the
14:58reason this works so well is that people
15:00are far more willing to engage with
15:02someone's perspective rather than a
15:05brand. So, a founder talking honestly
15:07about the problem they're solving, the
15:09mistakes they've made, and what they've
15:11learned, that is something that people
15:13are willing to engage with. That's what
15:14people share, what people screenshot,
15:17and what people come back to. And the
15:19best part, by the the they find your
15:20product, they already know who you are.
15:22They already trust your thinking, and
15:24the sale is already halfway done. So,
15:27for Searchable, our ideal customers are
15:30founders, marketing managers, and
15:31agencies. [music]
15:32So, every piece of content that I put
15:34out is built around problems that they
15:37are trying to solve, how they grow their
15:38brand, how to think about AI search, how
15:40to reach customers who are completely
15:43changing their behavior. And that means
15:45that every post, every video, every
15:47cheat sheet is designed to reach this
15:48audience and position Searchable as the
15:51perfect solution. The best part is that
15:52any organic content that you put out
15:54that performs well and drives people to
15:56your product, you can then boost later
15:58on with paid adverts. So, you're testing
16:00whether your paid ad strategy works
16:02before you're even running ads. So, the
16:04second layer is GTM. You need to get
16:06emotion going where you're reaching out
16:08to people from your waitlist, and you're
16:10introducing your business to them.
16:11You're giving them a chance to book a
16:13free 30-minute consultation and a
16:15product demo with them. And these warm
16:17prospects who come to you through lead
16:19magnets and content, they'll be the
16:21lowest friction clients because you've
16:23already delivered them some value, and
16:24they've already committed [music] a
16:26certain level of trust to you. And once
16:28that's in motion, you can layer in cold
16:30outreach. And this has to be highly
16:32personalized, value-first, and aimed at
16:35prospects who are perfect for your
16:37business. So, instead of leading with,
16:39"Hey, would you like a demo?" you should
16:41offer to diagnose something for them.
16:43So, try this. "I've been looking at your
16:45business, and I think there are three
16:47clear opportunities around
16:49the problem that you solve. [music] I've
16:51dropped this into a document, but I'm
16:52happy to jump on a call and talk you
16:54through what I'm seeing." That starts a
16:56conversation, it adds value, and it gets
16:58them on the call. So, the final layer is
17:00AEO, or answer engine optimization. This
17:03is about making sure your brand appears
17:06in AI-driven search, not just in Google,
17:08but ChatGPT,
17:10Perplexity, and other platforms where
17:12your customers are now looking for
17:14answers. So, the shift happening right
17:16now is the entire reason that we
17:18launched Searchable. People are using AI
17:20to make purchasing decisions, to choose
17:23which tool to use, which agency to hire,
17:25and which product actually solves their
17:27problem. And the brands that are showing
17:29up in these answers [music]
17:30are building a compounding distribution
17:32advantage that is going to get harder
17:34and harder to close the longer that you
17:36wait. So, we've just talked about
17:37building an MVP, collecting feedback,
17:40and then how to drive traffic to your
17:41[music] business. However, as a startup,
17:44the most important thing is that you
17:45make the most of any traffic and
17:47interest that you get, and you turn that
17:50into interested customers. And that is
17:52why it's so important early on to create
17:54high-quality lead magnets. The
17:56mechanisms are really straightforward.
17:58You create something genuinely valuable,
18:00a tool, a report, a framework, a piece
18:02of data your audience can't access
18:05anywhere else. And you put a small
18:06exchange [music] in front of it, usually
18:09an email address or a phone number. And
18:11that exchange
18:12is the moment that you own the
18:13relationship. Every piece of
18:14communication that you send afterwards
18:16goes straight to them. It doesn't just
18:17get posted [music] to social media. For
18:19Searchable, we spent a week building
18:21what became our most important
18:23acquisition tool, the free AI visibility
18:26report. Anyone could enter their company
18:28name, and Searchable would scan their
18:30website, identify their competitors, and
18:32generate a report showing exactly how
18:35they were appearing or not appearing
18:37across the major AI platforms compared
18:39to their competitors. And that's how we
18:41engineered this value exchange. Our
18:43report was free to generate, but to read
18:46the full results of the report, you
18:47needed to create a Searchable account.
18:49And that was it. But, it was only one
18:51step, with no credit card, no
18:52commitment, just an account. But, what
18:55we received in return was nearly 50% of
18:58those free users converted to paid
19:01users, all because we gave them an
19:02incredibly valuable piece of
19:04information, an incredibly valuable tool
19:06up front for free. We've actually just
19:08released the latest version of
19:10Searchable's visibility report. And if
19:12you want to see how your company ranks
19:14on ChatGPT and other LLMs for free, the
19:17link is at the top of the description in
19:18the video.
19:22So, now we're onto the launch. So, in
19:24January, we opened the doors to
19:25Searchable publicly. And thanks to the
19:27work that we put in on all the different
19:29stages we talked about, we reached a
19:31million dollars of ARR in 60 days, which
19:34let us immediately hire more people into
19:36our product and engineering team and
19:38build out our broader marketing team. I
19:40believe if follows this playbook, finds
19:42a niche within an incredible natural
19:45growth market, and develops a product
19:47that people are very interested in, they
19:49will get similar inbound from investors.
19:51So, the key tip is build [music] in
19:53exciting space, share your success,
19:56force early traction, and investors will
19:58be interested. But, let me end with a
20:00few ways that we are still driving this
20:02growth because scaling customer
20:04acquisition beyond social is challenging
20:06and complicated. But, there are two
20:08focus areas that we're working on right
20:10now. Inbound inquiries, which we're
20:12generating through adverts, through
20:14organic social media content, through
20:15press, and also ranking in Google,
20:18traditional search, and appearing in AI
20:20search as well. We also launched the AI
20:23Search Accelerator, which is a five-week
20:25live training program for businesses who
20:27are looking to drive their presence in
20:28ChatGPT. It attracted hundreds of
20:31customers that we've now trained
20:33hands-on, which is both great for
20:35awareness, but also for stickiness
20:37within the product.
20:39The second is outbound, which is the
20:40process of locating the contact details
20:43for your target ICP and sending them
20:45messages to check your tool out, to book
20:47a demo, and to see how your business
20:49works. We use a combination of Apollo
20:51and LinkedIn to find the data, and then
20:53we use Clay to enrich it to give us the
20:55contact details, and then we send really
20:57high-quality personalized outbound about
20:59their business and how Searchable could
21:01help them. And that brings us to today,
21:03the beginning of June. We've been live
21:05now for 5 months and have managed to
21:07scale our business to beyond 3 million
21:09dollars of annual recurring revenue. We
21:11have a team of 18 people and we are now
21:13working on the next iteration of our
21:14product and scaling up our business.
21:16>> [music]
21:16>> The next version of Searchable is an
21:18autonomous AI agent that will actually
21:20help you and your business to strategize
21:23and execute [music] the work required to
21:25build your business and appear in AI
21:28search. It's been a hell of a ride and
21:30we're obviously [music] not stopping
21:31now. So, I hope you continue watching
21:33the journey. And if you've made it this
21:35far, you're more committed than 99% of
21:38people. So, I want to thank you and if
21:40there is anything you want me to cover
21:41from building a business to marketing,
21:43just comment below and I'll make sure
21:45that I cover that in the next one.

About this transcript

This page contains the full transcript of I Built An $85M AI Startup in 5 Months (here's my playbook) by Chris Donnelly, generated from the public captions YouTube serves with the video. The transcript has 4,281 words across 646 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to translate the transcript into 100+ languages.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, translation to 100+ languages, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.